Contamination of Water by Synthetic Polymer Tubes Gregor A. Junk,* Harry J. Svec, Ray D. Vick, and Michael J. Avery A m e s Laboratory-USAEC and Department of Chemistry, Iowa State University, A m e s , Iowa 50010
Organic contamination ranging from 1-5000 parts per billion (ppb) by weight was detected in the water which had flowed through tubes of polyethylene, polypropylene, black latex, six different formulations of polyvinylchloride, and a plastic garden hose. The contaminants in the effluent water were isolated by sorption on macroreticular resin beads contained in a small glass column. The sorbed organic compounds were then eluted with diethylether, the eluate was concentrated by evaporation, and the organic contaminants were separated and measured quantitatively by gas chromatography. Identifications of plasticizers and other polymer additives were made by combination gas chromatography-mass spectrometry. The described method is accurate and convenient for testing polymer tubes intended for use in situations which require flowing water or water solutions.
Data have recently appeared concerning the production ( I , 2 ) and use ( 2 ) of certain plasticizers and other additives used to formulate polymers with properties desirable for their use in medicine, transportation, home furnishings, construction, apparel manufacture, food packaging, beverage production, and milk processing. The total production of phthalic acid esters, which represent only one of several classes of additives, has been over 1 billion lb per year during the period 1967 to 1973 ( I , 2 ) . This amount of synthetic material repetitively distributed throughout the environment naturally causes some concern, especially in view of the many recent toxicity studies of additives (3-16). Phthalates have been found in plants (17), fungus ( I 8 ) , air ( I , 19), food (20, 21), milk (22, 23), soil (24-27), blood (3, 5, 28-32), lipoid solutions (3, 5, 29), oil (33), body tissue (34), and water (5, 9, 35-40). Thus, the ubiquity of phthalates as well as other additives is clearly established. Although some of these may be present in the environment because of natural processes, there is little doubt that the major share of the current distribution is due to contamination from various polymers. This report describes an accurate method for the determination of the organic contamination of water by various polymer tubes. The method accommodates test conditions similar to those encountered in the eventual use of the polymer tubes to transport water or aqueous solutions.
Experimental XAD-2 Resin. This macroreticular resin was obtained as 20-60 mesh beads from Rohm and Haas, 5000 Richmond St., Philadelphia, Pa. 19137. The resin was purified by sequential Soxhlet extractions with methanol, acetonitrile and diethyl ether (41) prior to its use in the extraction columns described later. Solvents. All solvents were either spectrograde or analytical grade. Whenever blank determinations suggested impurities detectable by flame ionization gas chromatography, these solvents were further purified by fractional distillation. Test Water. The distilled water used in the contamination tests was purified by passage through a column filter 1100
Environmental Science & Technology
containing 20 grams each of XAD-2 resin and activated charcoal. Water treated in this manner contained less than one part per billion (ppb) by weight of detectable organic material. This purified test water is referred to as “pure” water in this report. Synthetic Polymer Tubes. All polymer tubes used in this study were purchased from commercial suppliers in 1973. These suppliers are not identified because polymer formulations are frequently changed and the results of our tests suggest that all commercial polymer tubes contaminate the water passed through them. The tubes were either 1/z in. or 3/s in. i.d. and cut to 25-ft lengths for test purposes. Pretreatment consisted of flushing -25 liters of “pure” water through each tube immediately before each contamination test. Instrumental. A single column, Varian 1200 gas chromatograph equipped with a linear temperature programmer and a flame ionization detector (FID) was used to separate, detect, and quantify the contaminants isolated from the water samples. All extracts of the contaminants were chromatographed using 6 ft X l / s in. 0.d. stainless steel columns packed with 80-100 mesh AW-DMCS treated chromosorb W coated with 5% w/w OV-1 liquid phase. Carrier gas flow was 20 ml/min for all separations and each chromatogram was obtained by injecting 3.0 pl of the ether concentrate into an off-column injector at 250°C. After an initial hold a t 50°C for 1 min, the column was programmed to 250°C a t 15”C/min. The detector temperature was 250°C. The detector response was calibrated for quantification using an aliquot of a standard solution of indan, naphthalene, and acenaphthylene in ethyl ether. A Du Pont model 21-490-1 combination gas chromatograph-mass spectrometer (gc-ms) was used to obtain mass spectral data from which many of the chromatographically separated components were identified. The same chromatographic conditions used with the Varian 1200 gas chromatograph were employed with the Varian 1400 gas chromatograph interfaced to the Du Pont 21-490 mass spectrometer via an all-metal jet separator. The separator and connecting lines were held at 225°C and the ion source at 275°C. A Digital Equipment Corp. (DEC) PDP-12/40 minicomputer was interfaced to the gc-ms to assist in data logging, reduction, and manipulations for interpretation purposes. Updated MASH (mass spectrometer data handling) software as supplied by DEC was employed. This software allowed for calibration of the computer for mass marking and acquisition of approximately 250 mass spectra acquired at six second intervals for each gc-ms run. The mass range covered was from 26-600. Data manipulations, such as generation of total ion mass chromatograms, selected ion mass chromatograms, background subtraction, spectra averaging, and ion series tables, were used at the discretion of the operator on an interactive basis after data accumulation and transfer onto magnetic tape by the minicomputer. These programs were used to aid in the interpretation of each mass spectrum. A visual display on a cathode ray tube was used for inspecting line diagrams of each mass spectrum to determine if a tabulation of the displayed spectrum via a Model 33 Teletypewriter was desired. In most cases, component identifications were suggested by a manual search of the Eight
Peak Index of Mass Spectra obtained from the British Information Sources, 845 Third Ave., New York, N.Y. 10022. When necessary, the DCRT/CIS Mass Spectral Search System, developed at the National Institutes of Health (42) and now available through the General Electric Mark I11 computer services (write to G.E. Information Services, 777 14th St., N.W., Washington, D.C., 20005), was also employed for identification purposes.
Tube Contamination Test Procedure Apparatus a.nd Quantification. A schematic diagram of the sampling apparatus is shown in Figure 1. All connections, adaptors, and valves are standard fittings except the upstream connection of the polymer tube to the valved intake manifold and the downstream connection to the glass extraction column which are both friction fit connections. The extraction columns are easily prepared by inserting a small silanized glass wool plug near the stopcock and pouring a methanol slurry of clean XAD-2 resin into the glass column. Then a second glass wool plug is inserted above the resin bed, and the column is capped to maintain the resin in a methanol-wetted condition. To test a polymer tube for contamination, the cap is removed from the column and the methanol is displaced with five 20-ml portions of “pure” water. The stopcock is closed when the last water wash level reaches the top of the upper glass wool plug. This procedure ensures complete displacement of the methanol and adequate water wetting of the resin. A 25-ft length of the polymer tube to be tested is then attached to one of the intake valves. The valve is opened and -25 liters of “pure” water are allowed to flow fipeely through the tube. This action flushes
n
I I
C
D
Figure 1. Sampling apparatus for contamination tests: (A) point of attachment of the combination XAD-2 and charcoal filter for purifying the distilled water. The arrow indicates the direction of water flow: (B) valved inlet manifold for simultaneous testing of several polymer tubes: (C) polymer under test: (D) glass extraction column containing 2 grams of clean 40-60 mesh XAD-2 resin: (E) glass wool plugs: (F) effluent stopcock and metering device for measuring total water flow through the polymer tube
the tube prior to the test and removes air bubbles which can cause undesirable turbulence during the test run. When the tube is free of air bubbles, the water flow is decreased and the downstream end of the tube is attached to the extraction column and the stopcock is opened immediately. The intake valve is adjusted to achieve a flow of 60 ml/min through the tube and extraction column. The simple expediency of checking the effluent water flow periodically with a graduated cylinder and a stopwatch is sufficiently accurate for calculating the total volume flow during a given test period. When the desired amount of water has passed through the tube, the intake valve and effluent stopcock are closed, and the extraction column is removed from the polymer tube. The procedure for eluting and quantifying the organic compounds sorbed on the resin has been described in detail by Junk et al. (41). Briefly this procedure is as follows: open the stopcock to drain most of the remaining water from the extraction column; add 25 ml of ethyl ether and allow 5-10 min for equilibration; open the stopcock and collect the ether eluate in a 20 X 150-mm test tube; plunge the test tube into liquid N2 for -20 sec to freeze out the 0.5-1.0 ml of residual water; decant the cold ether eluate into a modified Kuderna-Danish evaporator vessel; concentrate the ether solution to 0.5 ml; and use a 3.0-pl aliquot of this concentrate for gas chromatographic analysis employing the conditions described in the instrumental section of this report. The chromatogram peak areas are then measured and the amount of contamination due to each separated component is calculated using the equation,
where PL = peak area of the ith chromatographic peak in cm2, S = detector response factor in pg/cm2, VI = volume of ether concentrate in p1, V, = injected aliquot of ether concentrate in pl, Vu = volume of water passed through the polymer tube in liters, and ppb, = parts per billion by weight of the ith component in the water. The total contamination equals 2 ppb, where n 1q1 equals the number of components detected. The reproducibility is generally &lo%. Blank determinations are made using the same volume of water and procedure, but deleting the polymer tube and connecting the extraction column directly to the intake valve. Identifications. Since the gc-ms results are used solely for identification purposes, the normal procedure is to concentrate the ether eluate further from 0.5 ml to -0.05 ml using a stream of N2 gas or free evaporation. This additional concentration facilitates the gc-ms operation and although some losses of the more volatile components occur, no difficulty is encountered in relating the peaks observed on the conventional FID chromatogram to those observed on the gc-ms total ion monitor (TIM) chromatogram. Aliquots of 1.5 p1 are injected for the gc-ms identification experiments using the chromatographic-mass spectrometric conditions described in the instrumental section. Preliminary identifications were made from the computer manipulated mass spectral data. These were usually confirmed by matching both the gas chromatographic retention times and the mass spectral fragmentation patterns of the unknowns with those obtained for authentic samples. When necessary, additional confirmatory evidence for the identifications was obtained using other instrumental analyses of fraction-collected components. Volume8, Number 13, December 1974
1101
Results and Dkcussion Separation and Quantification. Tracings of the FID chromatograms, accumulated on the Varian 1200 gc (see Instruments section) for the various polymer tubes tested, are shown in Figures 2-4. The response is scaled for each chromatogram to make possible a rapid visual comparison of the number and the intensity distribution of contaminants. The total contamination, listed in the figure leg-
ends, varies from a low of 1 ppb for polyethylene tubes to a high of 5000 ppb for Food-Beverage Grade PVC tubes. To aid in the comparison of chromatograms, the chromatographic conditions and the time scales are identical in each figure. These comparisons reveal that a number of the contaminants have similar gc retention volumes. Also, from 50-90% of the total contamination normally involves four to six contaminants. The complexity of the chromatograms and the relatively
FOOD AND U
20
I6
~
12
8
4
0
0
-MINUTES-
-
-MINUTES-
16
20 MINUTES-
PROC.
, 20
jll ,'A,
" l6
I2 8 -MINu+ES-
4
I2 -MINUTES-
1" 8
4
HOSPITAL I7
9
sI E ) E V E R A G E
9 4
, I
0
,
0
Figure 2. Chromatograms from tests of polyvinyl chloride (PVC) tubes-25 ft X '/* in. i.d.: general chemical grade (11 p p b ) ; laboratory grade (4 p p b ) ; processed milk grade (6 p p b ) ; food-beverage grade (5000 p p b ) ; FDA-USDA approved grade (9 p p b ) ; hospital-surgical grade (6 p p b ) . Vinyl chloride monomer, if present, is not detected because it is masked by the large solvent peak 1102
Environmental Science & Technology
high level of contamination from all polymers except polypropylene and polyethylene are due primarily to the use of large amounts of plasticizers and other additives in these polymers. Also, impurities in the starting materials which become trapped in the polymerization process contribute to the observed contamination. For polyethylene and polypropylene, the contaminants present in the water are probably due solely to impurities in the starting materials and nonplasticizer additives such as stabilizers. The changing mass spectral fragmentation patterns observed across many of the apparent monocomponent gc peaks in Figures 2-4 suggested that the contamination mixtures were much more complex than indicated by the packed column chromatography employed for these separations. Preliminary work using support-coated open tubular (SCOT) columns reveals that as many as 100 components may be present in the water passed through some of the tubes. Identifkations. Although identifications of several contaminants present in minor amounts have been made, the primary purpose of this report is to focus on the major contaminants. To this end, detailed identification and quantification data compilations as illustrated by Table I for General Chemical Grade PVC were used to prepare a
summary of the five most intense gc peaks observed for each polymer tube. This summary is given in Table I1 where the gc peak numbers refer to those given in Figures 2-4. From Table 11, the phthalate esters appear most frequently among the five most dominant contaminants. Relatively large amounts of 2-ethylhexanol are present in the water from several of the polymer tubes and this may be due to hydrolysis of the phthalate ester and/or impurities present in the DEHP plasticizer employed in the production of the tubes. This observation, along with that of contaminants such as phthalic anhydride, lauric acid, and stearic acid used as additives only in the chemically combined state, indicates that the identification of contaminants from plastics is complicated by chemical transformation in addition to those problems associated with the use of proprietary additives. The total contamination profile for the various polymer tubes is summarized in Table I11 where both the major and minor contaminants in the water passed through the tubes are listed. An X designates that positive evidence for the contaminant exists. The absence of an X should not be construed as conclusive negative evidence. For example, the level may be below our detection limit, or a
POLYPROPYLENE n
7 2
-
5
20 12 -MINUTES-
20
8
4
12
-
I6
MINUTES
0
5
8
/I
0
4
POLY ETHYL ENE
w m z
9
B
E K
lif' I .A
20
16
-
12
8 MINUTES-
4
0
Figure 3. Chromatograms from tests of plastic garden hose and black latex tubes-25 ft X '/z in. i.d.: garden hose (170 ppb), black latex (26 ppb)
20
16
12
-MINUTES-
8
4
0
Figure 4. Chromatograms from tests of polyethylene and polypropylene tubes-25 f t X 3b in. i.d.: polyethylene (1 ppb), polypropylene (4 ppb)
Volume8, Number 13, December 1974
1103
particular component cannot be identified positively because of unresolved gc peaks, or a major contaminant masks a minor component. The identification results for the polyethylene and polypropylene tubes are not included here since these results are not yet definitive enough to be discussed. Static vs. Flowing Water. Interpretation of several experimental observations suggested that the amount of contamination from various polymer tubes may be related to the linear velocity of the water flow through the tube. An increase in the amount of contamination occurred when: the water flow rate was increased from 20 to 60 ml/min to reduce the sampling time, sharp bends existed in the polymer tube, and air bubbles were present in the tube during part or all of the sampling time. These observations, coupled with the implication that agitation of aqueous solutions may increase the contamination from PVC containers (22, 29, 34), suggested that linear velocity and contamination were indeed related. To test the validity of this proposal, “pure” water was passed through a
6-ft length of 1/2-in. i.d. Food-Beverage Grade PVC for 48 hr at 60 ml/min. The total measured contamination of the effluent water, a t a calculated linear velocity of 46 cm/ min, was 0.6 ppm. Then a 3-ft length of 3/s-in. glass rod was inserted into the center of the 6-ft PVC tube. The head pressure was increased to achieve the same flow of 60 ml/min and this flow was continued for 48 hr. Under these conditions, the linear velocity across the restricted 3-ft section was 105 cm/min, and the contamination in the effluent water increased to 1.6 ppm. This represents a dramatic enhancement of the contamination and suggests that it is directly related to linear velocity when the amount of water sampled, the sampling time, and the exposed polymer surface area are held constant. Previous investigations by others (22, 43, 44) have related the contamination mechanism to the migration of water into the polymer, to the diffusion of plasticizers and other additives through the polymer network, and to the dissolution rate of the additives. Our results suggest that an erosion mechanism occurring a t the polymer-water in-
Table I. identification and Quantification Data for Contaminants in Water Passed Through a 1/dn. I.D. Tube of General Chemical Grade PVC Peak Fa
Mass spectral datab
1 2 3 4 5 6 7
a
I CL CL CL CL HM HM HM
4.50 5.45 5.75 6.00 6.40
13 14 15
Concentration, ppbe
N N
6.90 7.30 7.80 8.40 9.70 10.00
CM I CL CL
Rt(k)d
3.60 3.95 4.25 5.40 N N
3.80 4.00
9 10 11
12
Gc data
Rt (unknp
yo of total concn
Identification
0.66
6.14
4.62 1.72 0.24 0.30 0.54 0.35 0.16
43.01 16.02 2.23 2.79 5.02 3.25 1.48 1.67
B u tylch loioacetate 2-Ethylhexanol o-Cresol Naphthalene p-Ethylphenol Isopropyl subs. phenol Methyl-ethyl subs. phenol Methyl-ethyl subs. phenol Unidentified Unidentified Unidentified Diethylphthalate Butyloctylfumarate Diisobutylphthalate Dibutyl p h t ha la te
0.18
11.50
9.60 N 11.80
12.50
12.40
1.02
0.11 0.08 1.45 0.08 0.05
0.74
13.49 0.74 0.46
a See Figure 2. b CM = a complete match of m s data with an authentic sample of the ICI-Aldermaston index, CL = close match. H M = match Of four largest peaks; I = no match and fragmentation pattern interpretation required. Retention time of unknown gc peak. d detention t i m e Of authentic sample. N indicates t h a t an authentic sample check was not made. e Concentration of the contaminant in the water in parts per billion by weight.
Table II. Total Contamination and Identifications of Major Contaminants in Water Passed Through Various 25-Ft Lengths of 3 / 8 - 1 / ? In. I.D. Polymer Tubes at 60 MI/Min Polymeric t u b e
Total contamination, p p b
General chemical
11
Laboratory grade PVC Processed milk
PVC
ldentificationsa Major contaminantsjgc peak n0.b
o-Cresol/3
Nap h t hale ne/4
BOF/13
2-Ethyl hexanol/2
4
2-Ethyl hexanol/l
Stearic acidll2
DIBP/9
BOF/6
6
2.Ethyl hexanol/2 DlBP/9
p-Nonyl phenol/lO
DEHP/11
methylphenol/8 3,4-Dimethoxy acetop henone/4 DEP/B
DEP/6
DEHP/15
EGEP/11
3,4-Dimethoxy acetop he none/5 2-Ethyl hexanol/4 Lauric acid/9
Unkn./4
DIBP/8
DBP/9
DEHP/17
DIBP/11
Phthalic anhydride/
Alcoholc/2 Decyl amine/2
Alcoholc/3 Unkn./8
Alcoholc/4 I onol/4
DBM/7 DAP/9
Unkn./l Unkn./2
Unkn./2 Unkn./4
U n kn./7 U n k n ./6
Unkn./5 Unkn./lO
DIBP/lO Diisopropyl p-xylene/3 U nkn./8 U nkn./l4
2-Ethyl-l-
PVC . .-
Food-beverage .PVC .FDA-USDA PVC Hospital-surgical PVC Garden hose Black latex Polyethylened Polypropylened 0
5000 9
6 170 26 1 2
DIBP/9
BGBP/13 2-Ethyl hexanol/2
7
Abbreviations used for common contaminants: BOF = b u t loctylfumarate; DlBP = diisobutyl hthalate: DEHP = di-2-eth Ihexylphthalate; DEP
= diethylphthalate; BGBP = butylglycolylbutylphthalate; EGEP = ethylglycolylethylphthalate’ g B P = dibutylphthalate. D g M = dibut Imaleate: DAP = diamylphthalate. b Refer to appropriate chromatograms in Figures 2-11. C Mass spectra 2nd retention times used to tentatively ldkntify this unresolved gc peakasa mixture ofalcohols.
1104
d
Insufficientwatersampled tomakeany positive identifications.
Environmental Science & Technology
terface should he included as a probable significant factor in the contamination mechanism. This relationship between contamination and the linear velocity of the water flowing through the polymer tube is currently being more fully investigated. Preliminary results show that the contamination of the water will increase as the flow rate is increased for a given diameter tube. Other factors being constant, larger diameter tubes and lower head pressures are recommended to minimize the contamination. An additional conclusion is that contamination levels, based on static water tests, cannot be extrapolated to flowing water conditions. Contamination vs. Usage. The possible decrease in the level of contamination after a polymer tube has been used for transporting large quantities of water was investigated. When 1000-liter volumes of “pure” water were used to flush Laboratory and Hospital-Surgical Grade PVC tubes, the amount of contamination measured immediately after the flushing action was reduced by a significant factor of -3 over that measured when only 25-liter volumes of “pure” water were employed for flushing. Subsequent flushings of the same tubes with as high as 3000 liters of water failed to decrease significantly the contamination. These results show that although initial flushing of PVC tubes with a layge quantity of water has a salutary effect in lowering contamination, extensive washing serves no useful purpose. The amount of plasticizers in the highly flexible PVC tubes is approximately 40% by weight and this represents im almost inexhaustible supply of contam-
ination. Should this supply eventually become appreciably depleted, the contamination will decrease, but the tubing will also have lost its flexibility and strength and will undoubtedly need to be replaced, thus causing a return to the original high level of contamination. For the relatively rigid polyethylene and polypropylene tubes the source of the contamination may not be inexhaustible, and these tubes may eventually become noncontaminating. Long-range tests of these tubes and various cleaning procedures are currently in progress.
Conclusions The method described here for testing the contamination of water by various polymer tubes is sensitive to amounts corresponding to < 1 ppb for most organic materials. Contamination tests are made, conveniently using flow conditions which duplicate or closely approximate the use of the polymer tube to transport water or other aqueous solutions. These tests can be made in most analytical laboratories and the results used to aid in the selection of polymer tubes. Since the critical apparatus in the method is a very simple and portable extraction column, accurate on-site sampling of a production solution is possible by connecting an extraction column parallel to a production stream until the desired volume is sampled. The extraction column may then be transported readily to the laboratory where the remainder of the scheme is completed at the convenience of the analyst. The contamination tests reported here were made using
~
~
~
Table Ill. List of Contaminants in Water and Polymer Tubes Which Caused This Contamination Tube abbreviationsa -.
GC
LEI
PM
FB
FU
HS
GH
2- Ethyl hexa no1 C8 satd. alcohol U n k n . alcoholij
X
X
X
X X
X
X
X
o-Cresol o-Ethylphenol p-Ethylphenol 2-Et hyl-4-methylp h e no1
X X X X
Contaminant
X X
X
X
X
X
X
X
-
X X X X
X
X
X
X X X
X X X
X X X
X X X X X X X
X X
X X X
-
Suspected stabilizer X
X
Probably octanol Not present in PVC Suspected stabilizer Same Same Same Same Same Same Same Same Highly toxic Common impurity
X
Dimethyl phthalate
Diamylphthala te Di-2-ethylhexyl p h t h a late
X
X
Butylglycolylbutylphthalate
Diethylphthalate Dibutyl p h t ha la te Di isob u tyl p h t h al ate
X X X
3,4-Dimethoxyacetophenone
Decylamine Stearic acid Lauric acid Myristic acid Ph t h a I ic a n h ytl ri d e Butyl benzoate Butyloctylfurnarate Di b u ty I m a I e a te Ethylglycolylethylphthalate
X X X
X
Comment
I mpurity or transformation*
X
2,4-Di-t-butyl-6.methylphenol 2,6-Di-t-butyl-4.methylphenol
p-Nonylphenol p -Dodecylpheiiol p-t-Butylphenol Ethyl acetate Butylchloroacetate Naphthalene 2-Ethylhexana I
BL
X X X X X
X
X
-
Transformation of Z n salt Same Same Impurity or transformation Known plasticizer Suspected plasticizer Same Same Known plasticizer Same Same Same Same
X
Same Same
Abbreviations are: GC = eneral chemical PVC. LB = laborator PVC, PM = processed milk PVC. FB = food-beverage PVC, FU = FDA-USDA PVC’ HS = hospital-surgical kVC. GH = garden hoke plastic: E L = {lack latex tube.,)( indicates t h a t {he contaminant was found’in the designated tube‘. bHere impurityreferstomatkrialsfound insolvents,startingchemicals,andaddltivesusedin the polymerization process. Q
~ - - _ _ _
Volume 8 , Number 13, December
1974
1105
distilled water because these contamination results are considered to represent minimum values. The presence of minerals, inorganic acids or bases, and lipoid materials in the water would all result in either the same or an increased level of contamination. In the presence of large amounts of minerals and inorganic acids and bases, the method has been shown (41) to be accurate for all neutral organics and therefore is not limited to distilled water tests. Whether the procedure is useful for beverage testing where lipoid materials are present has not yet been ascertained. The complexity of the chromatograms for all the tests of polymer tubes reported here suggests that conclusions concerning toxicity based on the tests of single pure components must be tempered with the knowledge that plasticizers and other additives almost always occur as a complex mixture. Several conclusions, based on the results reported here, suggest that risk/benefit ratios associated with the widespread use of many polymers may not be nearly so favorable as suggested by sole consideration of the reported low level of contamination. First, the contamination profiles represent a very complex mixture of components, many of which are not identified and may be toxic. Second, some of the identified components are obviously toxic, for example butylchloroacetate. Third, the presence of myristic, stearic, and palmitic acids in some polymer tests suggests that some toxic metals may also be present. Fourth, contamination level predictions based on static water tests of polymers are usually not valid. When these conclusions are coupled with the observations of other investigators, the case for reassessment of the risk/benefit ratio becomes even more convincing. Autian (3) discusses the documented increase in contamination of PVC‘containers when the contacting solution becomes more lipid; several investigators have established the rapid biomagnification of many plasticizers and other additives (9, 15, 37, 39); while little is known about cumulative toxicity and the synergistic effect of mixtures ( 3 ) , some investigators (3, 12, 35) have shown these to be appreciable; and finally, recent studies (26, 27) suggest that plasticizers may be readily transported as water-soluble complexes with natural humic materials. Our results and the summary outlined above suggest that the expressed concern of some scientists (3, 8, 32, 37, 39) is certainly valid. Unfortunately, definition of the true problem and the development of an adequate solution are severely complicated by the continuously changing and expanding list of allowable additives in the manufacture of polymer tubes and containers.
Acknowledgment The authors thank Neil G. Johansen, Perkin Elmer Corp., for providing high resolution SCOT column separations of two mixtures. The facilities for this research were generously provided by the Ames Laboratory of the USAEC and Iowa State University Energy and Mineral Resources Research Institute, Ames, Iowa 50010. Literature Cited (1) Gross, F. C., Colony, J . A., Enuiron. Health Perspec., 1 (3), 37 (1973).
1106
Environmental Science & Technology
(2) Graham, P. R., ibid., 1 ( 4 , p 3. (3) Autian, J., ibid. (4) Calif. Med., 112 (2), 43 (1970). (5) Buadze, 0. P., Gegeshidge, G. A., Nazin, A. N., Shmyreva, T . A., J. Polym. Sei., 33,349 (1971). (6) Guess, W. L., Haberman, S.,J. Biomed. Mater. Res., 2, 313 (1968). (7) Haberman, S., Guess, W. L., Rowan, D. F., Bowman, R. O., Bower, R. K., SOC.Plast. Eng. J., 34,62 (1968). (8) Mayer, F. L., Chem. Eng. News, 49 (45), 8 (1971). (9) Mayer, F. L., Stalling, D. L. Johnson, J . L., “Phthalate Esters; An Environmental Contaminant,” rep from Fish-Pesticide Research Lab., Bur. Sport Fisheries and Wildlife, U. S.Dept. Interior, Columbia, Mo., 1971. (10) Meyler, F. L., Willebrand, A. F., Durrer, D., Circ. Res., 8, 44 ( 1960). (11) Neerguard, J., Nielsen, B., Faurby, B., Christensen, D. H., Nielsen, 0. F., Scand. J . Urol. Nephrol., 5 , 141 (1971). (12) Nematollaki, J., Guess, W. L., Autian, J., J. Pharm. Sei., 56,1446 (1967). (13) Shaffer, C. B., Carpenter, C. P., Smyth, H. F., J . Ind. Hyg. Tonicol., 27,130 (1945). (14) Stetson, J . B., Guess, W. L., Anesthesiol., 33,635 (1970). (15) Marx, J. L., Science, 178,46 (1972). (16) Shea, K. P., Enuiron., 13 (a), 2 (1971). (17) Hayashi, S., Asakawa, Y., Ishida, T., Matsuura, T . , Tetrahedron Lett., 50,5061 (1967). (18) Sugiyama, N., Kashima, C., Yamamoto, M., Sugaya, T . , Mohri, R., Bull. Chem. Soc. Japan, 39,1573 (1966). (19) Blumer. M.. Contam. Contr.. 4,13 (1965). (20) Shibko, S. I., Blumenthal, H., Enuiron. Health Perspec., 1 (3),131 (1973). (21) Wandel, M., Tengler, H., Debut. Lebensm.-Rundsch., 59, 326 1963). (22) Wildbrett, G., Enuiron. Health Perspec., 1 (3), 29 (1973). (23) Cerbulis, J . , Ard, J. S., J. Ass. Offic. Anal. Chem., 50, 646 ( 1967). (24) Cifrulak, S.D., Soil Sci., 107,63 (1969). (25) Bauman. A. J.. Cameron. R. E.. Kritchevskv. G.. Rouser. G., Lipids, 2,85 (1967). (26) Ogner, G., Schnitzer, M., Science, 170,317 (1970). (27) Matsuda, K., Schnitzer, M., Bull. Enuiron. Contam. Toxicol., 6,200 (1971). (28) Guess. W. L.. Jacob.. J.., Autian. J.. Drug Intel.., 1,. 120 (1967). (29) Jaeger, R. J., Rubin, R. J . , Science, 170,460 (1970). (30) Marcel, Y. L., Noel, S. P., Lancet, 5,35 (1970). (31) Rubin, R. J., ibid., 10,965 (1972). (32) Trimble, A . S., Goldman, B. S., Yao, Y. K., Kovats, L. K., Bigelow, W. G., Surgery, 59,857 (1966). (33) Phillips, H . F., Berger, I. A , , Geochim. Cosmochim. Acta, 15,51 (1958). (34) Jaeger, R. J., Rubin, R. J., Enuiron. Health Perspec., 1 (3), 95 (1973). (35) Stalling, D. L., Hogan, J . W., Johnson, J. L., ibid., p 159. (36) Hites, R. A , , ibid., p 17. (37) Corcoran, E . F., ibid., p 159. (38) Shaposhnik, S.,Kodner, M . S., Chapurin, V. I., Borisenko, G. M., Tr. Mold. Nauch.-Issled. Inst. Pishch Prom., No. 10, 132 (1970). (39) Metcalf, R. L., Booth, G. M., Schuth, C. K., Hansen, D. J., Lu, P. Y., Enuiron. HealthPerspec., 11 (4), 27 (1973). (40) Pastorelli, L., Chiavari, G., Ann. Chim. (Rome), 61, 311 (1971). (41) Junk, G. A,, Richard, J . J., Grieser, M . D., Witiak, D., Witiak, J. L., Arguello, M . D., Vick, R., Svec, H . J., Fritz, J . S., Calder, G. V., J. Chromatogr.. 99, 745 (1974). (42) Heller, S.R.,Anal. Chem., 44, 1951 (1972). (43) Wildbrett, G., Evers, K., Kiermeier, F., Fette, Seifen, Anstrichm., 71,330 (1969). (44) Quackenbos, H. M., Ind. Eng. Chem., 46, ( 2 ) . 1335 (1954). 1
Receiued for review May 13, 1974. Accepted September 13, 1974. Work partially supported by the National Science Foundation, Grant No. GP-33526X.
